Acta Psychologica 124 (2007) 370–381
www.elsevier.com/locate/actpsy
The effect of orientation on number
word processing
Roi Cohen Kadosh *, Avishai Henik, Orly Rubinsten
Department of Behavioral Sciences, Zlotowski Center for Neuroscience,
Ben-Gurion University of the Negev, P.O. Box 653, Beer-Sheva, Israel
Received 3 November 2005; received in revised form 20 April 2006; accepted 26 April 2006
Available online 14 June 2006
Abstract
Besner and Coltheart [Besner, D., & Coltheart, M. (1979). Ideographic and alphabetic processing
in skilled reading of English. Neuropsychologia, 17, 467–472] found a size congruity effect for Arabic
numbers but not for number words. They proposed that Arabic numbers and number words are processed in different ways. However, in their study orientation of the stimuli and notation were confounded. In the present study, it is found that orientation of number words affects numerical
processing. Orientation modulates both the size congruity effect and the distance effect; horizontal
presentation produces similar results to those produced by Arabic numbers whereas vertical orientation produces different results. Accordingly, it is proposed that our cognitive system is endowed
with two different mechanisms for numerical processing; one relies on a visual-spatial code and
the other on a verbal code.
2006 Published by Elsevier B.V.
PsycINFO classification: 2300; 2340; 2346
Keywords: Cognitive processes; Attention
*
Corresponding author. Tel.: +972 8 6477209; fax: +972 8 6472072.
E-mail address: [email protected] (R. Cohen Kadosh).
0001-6918/$ - see front matter 2006 Published by Elsevier B.V.
doi:10.1016/j.actpsy.2006.04.005
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
371
1. Introduction
Studies in the field of number processing documented two well-known effects. The
first, referred to as the numerical distance effect, was first reported by Moyer and Landauer (1967). As the name implies, the distance between two numbers influences the comparison of the stimuli; the larger the distance, the shorter the reaction time (RT). The
second effect is called the size congruity effect (Henik & Tzelgov, 1982) and is a
Stroop-like phenomenon. When a stimulus has two dimensions but only one has to be
attended while the other has to be ignored, participants process the irrelevant dimension
unintentionally. Interference occurs when the two dimensions are incongruent and facilitation occurs when the two dimensions are congruent. For example, in number comparison, participants have to relate to the numerical value and ignore the physical size. In
the congruent condition one of the two digits is larger in both dimensions (e.g., 2 4).
In the incongruent condition one of the digits appears larger in the relevant dimension,
whereas the other appears larger in the irrelevant dimension (e.g., 2 4). In the neutral
condition there is no difference in the irrelevant dimension (e.g., 2 4). Facilitation is
observed when the RT to the congruent trials is faster than that of the neutral trials.
Interference is observed when the RT to the incongruent trials is slower than that of neutral trials.
The size congruity effect for the number comparison was first observed by Besner and
Coltheart (1979). Since then, it has been shown that this effect is found not only when the
numerical value is the relevant dimension and the physical dimension must be ignored (as
shown by Besner and Coltheart), but also in the reverse task, when the physical dimension is relevant and the numerical dimension is irrelevant. This implies that not only
physical size but also numerical value is processed automatically (e.g., Henik & Tzelgov,
1982).
In addition, an interaction between the size congruity effect and the distance effect was
found in many studies (e.g., Cohen Kadosh & Henik, 2006a, 2006c; Cohen Kadosh,
Henik, & Rubinsten, 2005a; Schwarz & Ischebeck, 2003), that is, as the numerical distance
between stimuli in the numerical task increases, size congruity decreases. It has been suggested that numbers are processed serially (e.g., Dehaene, 1996; Schwarz & Ischebeck,
2000). Under serial processing, according to the additive factor method (Sternberg,
1969), an interaction between two factors occurs if processing of both factors is carried
out at a common stage of processing. Hence, the interaction between numerical distance
and size congruity suggests that processing of both factors is carried out at the comparison
stage. Indeed, it was found that a shared area in the intraparietal sulcus (IPS) is activated
for the numerical distance effect and for the size congruity effect (Kaufmann et al., 2005;
Pinel, Piazza, Le Bihan, & Dehaene, 2004).
Neuroimaging studies reported shared processing of number words (e.g., EIGHT) and
Arabic numbers (e.g., 8) at the comparison stage (e.g., Dehaene, 1996; Naccache & Dehaene, 2001; Pinel, Dehaene, Rivie‘re, & Le Bihan, 2001; Pinel et al., 1999). These results
support the idea that numbers are abstractly represented on a mental number line. Additional support for the mental number line comes from a finding that participants respond
faster to small numbers with their left hand and to large numbers with their right hand.
This effect was first shown by Dehaene, Bossini, and Giraux (1993) and labeled the
SNARC effect (spatial numerical association of response codes). Since then, it was
reported by other groups (e.g., Bachtold, Baumuller, & Brugger, 1998; Berch, Foley, Hill,
372
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
& Ryan, 1999; Fias, Brysbaert, Geypens, & d’Ydewalle, 1996; Fias & Fischer, 2004; Fias,
Lauwereyns, & Lammertyn, 2001).
In contrast to the idea of notation independent representation of numbers, Besner
and Coltheart (1979) found a size congruity effect with Arabic numbers (i.e., the irrelevant physical size affected numerical processing) but failed to find the same effect with
word notation (i.e., the irrelevant physical size did not affect numerical processing of
number words). They argued that these results indicated a distinction in the processing
of Arabic and number words. However, it is important to note that in their experiment
the two notations differed in orientation also. Namely, the Arabic notation appeared
horizontally (one stimulus on the left and the other on the right) whereas the word notation appeared vertically (one stimulus above the other stimulus), leading to the possibility that the difference in orientation produced the discrepancy in the results.
Accordingly, it is possible to argue that the vertical presentation of number words
was incompatible with the mental representation of numbers, which in turn, slowed
down processing of the number words so that by the time the numerical values were
present in the system, activation of the physical dimension dissipated and did not affect
numerical processing. In contrast with this idea, Foltz, Poltrok, and Potts (1984) argued
that the absence of the size congruity effect for number words in Besner and Coltheart’s
report was an artifact of their experimental design. They suggested that the repetition of
the same pairs of stimuli during the experiments eliminated the effect. When participants
encountered a repeated pair they could retrieve the decision reached previously with the
same pair (i.e., ‘‘the right number word was larger’’ or ‘‘the left number word was larger’’). This would allow for ignoring the irrelevant dimension in most trials in Besner
and Coltheart’s study. Note that Foltz et al.’s criticism did not refer to the sequential
effect, namely, the effect that a repeated stimulus in trial N had on the processing of
the identical stimulus in trial N + 1 (Schwarz & Ischebeck, 2000). Rather, they referred
to the effect of retrieval from the memory of a repeated stimulus, independent of the trials that preceded it.
Foltz et al. replicated Besner and Coltheart’s experiment and found a size congruity
effect for number words as well as for Arabic numbers. However, their experiment differed
from Besner and Coltheart’s experiment in two aspects. First, each pair appeared only
once, and second, their stimuli were presented only horizontally and not vertically. Therefore, one cannot be sure whether the size congruity effect for number words in Foltz et al.’s
study was due to the orientation of the stimuli presentation.
2. The current study
This experiment was carried out in order to investigate whether the size congruity effect
can be modulated by the orientation of the number word stimuli. Participants were presented with two separate blocks of trials and asked to judge which stimulus was numerically larger. In one block stimuli were presented horizontally and in another block stimuli
were presented vertically, which is similar to Besner and Coltheart (1979) experiment. In
contrast to the work done by Foltz et al. (1984), each pair appeared more than once in our
design.
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
373
3. Method
3.1. Participants
Seventeen university students (mean age 22.23, SD = 1.33) from Ben-Gurion University
of the Negev participated in the experiment for partial fulfillment of a course requirement.
All the participants had normal or corrected-to-normal vision, had no reading or mathematical deficits, and their mother tongue was Hebrew.
3.2. Stimuli
A computer display consisted of two number words in Hebrew which appeared at the
center of the screen. In the horizontal block the center-to-center distance between the two
number words was 10 cm. The participants sat 55 cm from the screen. The stimuli subtended a vertical visual angle of 0.7 or 0.9 and a horizontal visual angle of 1.7–5.4.
The same stimuli were used for the vertical and the horizontal blocks.
There were three types of pairs: congruent, neutral and incongruent. The number
words, one through nine excluding five, were used to produce three numerical distances:
1 (the number word pairs 1–2, 3–4, 6–7, or 8–9), 2 (the number word pairs 1–3, 2–4, 6–
8, or 7–9) or 5 (the number word pairs 1–6, 2–7, 3–8, or 4–9). Each number word was presented for an equal number of times for each distance. Stimuli were arranged in blocks of
trials (i.e., horizontal or vertical block) with each block composing 72 different stimuli that
were presented twice (a total of 144 trials in each block). Each number word and each
physical size appeared for an equal number of times on the left and the right. Each block
had nine different conditions; three numerical distances · 3 congruency conditions. Each
condition was represented by 16 trials (i.e., 4 number word pairs · 2 physical sizes · 2 sides
of the computer screen) in a given block.
Every experimental block was preceded by a block of 36 practice trials. This block was
similar to the experimental block with the exception that we used number words that were
different from those employed in the experimental blocks. For a numerical distance of 1
unit, the practice number words were made out of the pairs 4–5 and 5–6, and for numerical
distance of 2 units the number words were 3–5 and 5–7. In the case of the 5 unit numerical
distance we used the number word pairs 1–6 and 4–9, which were also used in the experiment. This was due to an error in the design of the experiment and we examine its effect in
the result section.
3.3. Procedure
The participant’s task was to decide which of two number words in a given display is
larger, with the term ‘‘larger’’ applying to the numerical value. Each participant took part
in one session composed of two different blocks. Participants were asked to respond as
quickly as possible but were asked to avoid errors, and attend only to the numerical values. The stimuli in each block were presented in a random order. Participants used the
index finger of both hands to indicate their choices by pressing one of two keys corresponding to the side of the display with the larger digit (i.e., vertically aligned keys for
374
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
vertical orientation, and horizontally aligned keys for the horizontal orientation). Each
trial began with an asterisk as a fixation point, which was presented for 300 ms at the center of a computer screen. Five hundred ms after the fixation point disappeared a pair of
number words appeared and remained in view until the participant pressed a key (but
not for longer than 5000 ms). A new stimulus appeared 1500 ms after the participant’s
response. The order of the two blocks was counterbalanced across individual participants.
3.4. Design
The variables manipulated were: orientation (vertical vs. horizontal), numerical distances (1, 2, or 5) and congruity (incongruent, neutral and congruent). Thus, we had a
2 · 3 · 3 factorial design.
4. Results
Error rates were generally low (3% and 3.5% for horizontal and vertical blocks, respectively). One participant was dropped from the analyses due to a high error percentage
(more than 10%). For the remaining 16 participants, for each condition mean RT was calculated for correct trials only. These mean values were subjected to a three-way analysis of
variance (ANOVA) with orientation, numerical distance and congruity as within-subject
factors.
There was a main effect of numerical distance [F(2, 30) = 42.56, MSE = 2000,
p < 0.001]. Participants responded faster to large numerical distances than to small ones,
with mean RTs of 844 ms, 807 ms, and 785 ms, for 1, 2, and 5 units distance, respectively.
There was also a significant effect of congruity [F(2, 30) = 31.27, MSE = 2384, p < 0.001],
with mean RTs of 838 ms, 816 ms and 783 ms for incongruent, neutral and congruent
pairs, respectively. The main effect for orientation was not significant [F < 1]. In addition,
two interactions were significant: orientation and numerical distance [F(2, 30) = 4.54,
MSE = 2009, p < 0.05], and orientation, numerical distance and congruity
[F(4, 60) = 3.29, MSE = 1592, p < 0.05]. The latter three-way interaction is presented in
Fig. 1.
It seems that there was a congruity effect regardless of orientation. The congruity effect
in the horizontal orientation seems to have both interference and facilitatory components
whereas the congruity effect in the vertical orientation was mostly based on facilitatory.
Moreover, the congruity effect seems to be modulated by numeric distance for horizontal
orientation but not for vertical orientation. Additional analyses of the three-way interaction which were conducted separately for horizontal and vertical orientation (Keppel,
1991), supported these conclusions.
4.1. Horizontal orientation
A distance effect [F(2, 30) = 39.20, MSE = 1775, p < 0.001], and a size congruity effect
indicated by the congruity simple main effect [F(2, 30) = 12.75, MSE = 3333, p < 0.001]
were obtained. For numerical distance we found a decrease of RT with distance
(845 ms, 808 ms, and 769 ms for the 1 unit, 2 units and 5 units distances, respectively).
The differences between 1 unit and 2 units, and between 2 units and 5 units were signifi-
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
375
A
Incongruent
Neutral
Congruent
1000
Reaction Time (ms)
Horizontal Orientation
900
800
700
Numerical
Distance
B
1
2
5
1000
Reaction Time (ms)
Vertical Orientation
900
800
700
Numerical
Distance
1
2
5
Fig. 1. Reaction time as a function of numerical distance and congruity under horizontal (A) and vertical
orientation (B). 95% confidence intervals are given for the triple interaction (Masson & Loftus, 2003).
cant, [F(1, 15) = 22.56, MSE = 1468, p < 0.001] and [F(1, 15) = 13.97, MSE = 2610,
p < 0.005], respectively. Examination of the components of the congruity effect revealed
a facilitatory component (congruent relative to neutral) of 22 ms [F(1, 15) = 6.13,
MSE = 1895, p < 0.05], and an interference of 37 ms [F(1, 15) = 10.64, MSE = 3065,
p < 0.005]. The simple interaction between congruity and numerical distance was significant [F(4, 60) = 3.08, MSE = 2134, p < 0.05], showing a decrease in the congruity effect
with the increase in numerical distance.
Examination of the facilitatory and interference components for each distance revealed
an interference component [F(1, 15) = 10.40, MSE = 4489, p < 0.01] but no facilitatory
[F(1, 15) = 1.26, MSE = 2381, ns] component for distance 1. In contrast, the reverse result
was obtained for distance 2, namely, a non-significant interference component [F < 1], and
a marginally significant facilitatory component [F(1, 15) = 4.44, MSE = 1481, p = 0.052].
376
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
B
A
60
Reaction Time (ms)
Distances
∗
50
1 vs. 2
∗
2 vs. 5
40
30
20
10
0
Horizontal
Orientation
Vertical
Orientation
Fig. 2. Reaction time differences for distance as a function of horizontal (A) and vertical (B) orientations in the
neutral condition (*p < 0.05).
For distance 5, both interference [F(1, 15) = 6.14, MSE = 1200, p < 0.05] and facilitatory
[F(1, 15) = 6.18, MSE = 416, p < 0.05] components were significant.
In order to examine the ‘‘pure’’ numerical distance (without the affect of physical size)
we analyzed the distance effect under the neutral condition separately (Fig. 2A). The
distance effect was significant [F(2, 30) = 12.67, MSE = 1368, p < 0.001]. Additional analyses revealed that the difference between 1 unit and 2 units was not significant [F < 1].
In contrast, the distance between 2 units and 5 units was significant [F(1, 15) = 15.91,
MSE = 1328, p < 0.001].
4.2. Vertical orientation
A distance effect [F(2, 30) = 11.03, MSE = 2233, p < 0.001] and a size congruity effect
indicated by the congruity simple main effect [F(2, 30) = 16.45, MSE = 2276, p < 0.001]
were obtained. For numerical distance we found a decrease of RT with distance
(843 ms, 807 ms, and 801 ms for the 1 unit, 2 units and 5 units distances, respectively).
The difference between 1 unit and 2 units was significant [F(1, 15) = 22.56, MSE = 1468,
p < 0.001], but the difference between 2 units and 5 units was not significant [F < 1]. Examination of the components of the congruity effect revealed that it was composed of only a
facilitatory component (congruent relative to neutral) of 44 ms [F(1, 15) = 12.34,
MSE = 3728, p < 0.005] whereas the interference component was insignificant [F < 1].
The simple interaction between numerical distance and congruity was also not significant [F < 1]. Since the interaction between distance and congruity was not significant,
we did not analyze the interference and facilitatory components for each distance
separately.
As in the horizontal orientation analysis, we analyzed the distance effect under the neutral condition (Fig. 2B). The distance effect was significant [F(2, 30) = 4.61, MSE = 2297,
p < 0.05]. The difference between 1 unit and 2 units was significant [F(1, 15) = 6.05,
MSE = 2355, p < 0.05]. In contrast, the distance between 2 units and 5 units was not significant [F < 1].
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
377
In order to examine the possibility as to whether the congruity effect was due to confounding of word length, a variable that Besner and Coltheart (1979) controlled, we ran
an additional analysis of the vertical orientation only. This analysis included only pairs
of words that had the same number of letters for the smaller numbers and the larger numbers. We excluded pairs with asymmetric length from the analyses (e.g., the pair six-eight,
in Hebrew:
, was excluded since there was a difference of three letters between
the two words). Similar to the results reported above, this analysis produced a significant
size congruity effect [F(2, 30) = 11.58, MSE = 4617, p < 0.001].
Since some of the 5 units pairs (i.e., the number word pairs 1–6 and 4–9) were used both
in the experimental and practice blocks, we analyzed only those pairs that were used during the experiment (i.e., the number word pairs of 2–7 and 3–8) in order to eliminate any
possibility of difference due to the practice effect. No significant difference, due to practice
was observed.
4.3. Error rates
Due to lack of variance in several conditions, a correlation analysis was conducted
between RTs and error rates. The results did not show any RT-accuracy trade-off
(r(144) = 0.31, p < 0.01 for horizontal block, and r(144) = 0.23, p < 0.05 for vertical block).
5. Discussion
Finding a size congruity effect in word notation seems to contradict Besner and Coltheart (1979) report. They reported a size congruity effect for Arabic numbers but not
for word notation. Accordingly, Besner and Coltheart concluded that Arabic numbers
and number words are processed differently. However, in their experiment the notation
was confounded with orientation of the stimuli (i.e., words were presented vertically
whereas Arabic numerals were presented horizontally). Hence, it is difficult to decide
whether the source of the discrepancy was difference in the notation or in the orientation.
In contrast to Besner and Coltheart’s results, a size congruity effect was obtained in our
study for vertical orientation as well as for horizontal orientation. The discrepancy
between our results and the null result reported by Besner and Coltheart might be due
to a stronger effect of orientation in their experiment (as can be indicated by the higher
error rate (above 10%) in their study in contrast to the current study). Indeed, in our
experiment the manipulation of orientation played a significant role. When we compare
the two orientations of the stimuli, it seems that the two effects that are of interest here
are modulated by orientation of the display: (1) The size congruity effect was composed
of a facilitatory component in the vertical orientation whereas it was composed of both
facilitatory and interference components in the horizontal orientation. (2) The same pattern of interaction between numerical distance and congruity as reported in other studies
(e.g., Cohen Kadosh & Henik, 2006a, 2006c; Cohen Kadosh et al., 2005a; Schwarz &
Ischebeck, 2003; Tzelgov, Meyer, & Henik, 1992), was observed in the horizontal orientation but was absent in the vertical orientation. (3) The ‘‘pure’’ distance effect (as reflected
by the neutral condition) showed modulation by orientations; in the horizontal orientation
only the difference between 2 units and 5 units was significant, whereas in the vertical orientation we found a mirror image (Fig. 2A and B), namely, only the difference between 1
and 2 units was significant.
378
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
All the differences between orientations that were mentioned above strengthen the idea
that the orientation of the stimuli affected the processing of the number words. We suggest
that when the stimuli appeared horizontally, they resembled the representation of the
mental number line and hence were less demanding on cognitive resources than in the
vertical orientation condition. In contrast, the vertical orientation of the stimuli, which
is incompatible with the representation of the mental number line, impacted on the congruity and distance effects. Consequently, in the horizontal orientation the distance and
congruity effects are implemented almost completely. It is important to note the opposite
pattern of the distance effect under different orientations (i.e., the difference between the
neutral conditions in the two orientations). A possible explanation for such a mirror image
might be the existence of two mechanisms for number comparison: (1) The more dominant
mechanism processes numbers using a visuospatial code. In this, code numbers are compared by representing them on a horizontal mental number line. Imaging and patient studies have pinpointed the IPS as the brain area that is dedicated to numerical comparisons
(for review see, Dehaene, Piazza, Pinel, & Cohen, 2003) by representing numbers on a
mental number line (Zorzi, Priftis, & Umilta, 2002). (2) An additional mechanism processes numbers via a verbal code. This suggestion is supported by fMRI studies that found
that numerical distance modulated brain activations in semantic language areas (superior
temporal sulcus (STS) and the middle temporal gyrus (MTG)) (Cohen Kadosh et al.,
2005a, 2005b; Pinel et al., 2001). Given the abstract nature of numerical information,
and the dependence between language and numbers (Wiese, 2003), it is plausible that such
temporal brain structures are also active during number processing. Bloch-David, Henik,
and Rafal (submitted for publication), and Cohen Kadosh et al. have proposed that numbers are connected to one another in similar way as representations of words are (e.g.,
DOG-CAT, TABLE-CHAIR). Thus, close numbers with a numerical distance of one unit
(e.g., ONE-TWO) are more difficult to compare because they are linked by stronger connections than numbers that are further apart (e.g., ONE-FOUR). The further apart the
numbers are, the less they are affected by the verbal code, due to the simple fact that connections between such numbers are negligible or even absent. When participants compare
numbers presented horizontally, the dominant visuospatial code is more active. Hence,
numbers that are further apart (e.g., numerical distance of 5 units) are processed faster
than a closer numbers (i.e., numerical distance of one and two units). However, in the case
of incompatible presentation of numbers (e.g., vertical presentation), the close numbers
can still be compared in an efficient way due to the verbal code, and the numerical distance
effect will still be observed. The existence of two mechanisms for numerical comparison
also supplies an explanation for the observable numerical distance effect after damage
to visuospatial abilities (e.g., Zorzi et al., 2002) or to the left and right IPS (Bloch-David
et al., submitted for publication).
The lack of interaction between congruity and the distance effect in the vertical orientation, which stands in contrast with horizontal orientation, also supports the possible
existence of two mechanisms for number comparison: The interaction between numerical
distance and size congruity suggests that processing of both factors is carried out at the
comparison stage. As was mentioned in Section 1, numbers are processed serially (Dehaene, 1996; Schwarz & Ischebeck, 2000). Based on the additive factor method (Sternberg,
1969) the interaction in the horizontal presentation between distance and congruity suggests that the processing is probably carried out at the comparison stage. This hypothesis
is supported by Kaufmann et al. (2005) and Pinel et al. (2004) who found that the IPS is
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
379
activated by the numerical distance effect and by the size congruity effect. The IPS is
known to play a role not only in numerical processing but also in visuospatial representation (Corbetta & Shulman, 2002; Hubbard, Piazza, Pinel, & Dehaene, 2005). Hence, it is
possible that with horizontal orientation the IPS is involved in processing the physical size
of the displayed numbers as well as their numerical value via a visuospatial code. Under
vertical presentation numerical distance and size congruity were additive. In this case it
is possible that numerical value is processed via a verbal code, probably by the temporal
structure (Cohen Kadosh et al., 2005b; Pinel et al., 2001), and physical size is processed
by the IPS (Cohen Kadosh et al., 2005b; Fias, Lammertyn, Reynvoet, Dupont, & Orban,
2003; Pinel et al., 2004).
The current idea that the overlap between brain structures can modulate chronometric
data is not new. It has been suggested that the degree of interference between relevant and
irrelevant information stems from the overlap in neural activity/brain structures (Cohen
Kadosh & Henik, 2006a, 2006b, 2006c; Fias et al., 2001; Kinsbourne & Hicks, 1978; Lammertyn, Fias, & Lauwereyns, 2002; Posner, Sandson, Dhawan, & Shulman, 1990). For
example, Fias et al. (2001) and Lammertyn et al. (2002) showed that irrelevant and relevant information which are processed by a shared brain area, as in the case of orientation
and representation of numbers, tend to interfere with one another.
6. Conclusions
The current study showed that the orientation of the presented stimuli can modulate the
distance and the size congruity effects. However, in the current study we used number
word notation. It is an open question whether such an effect will also be found with other
notations such as Arabic numbers, or roman numerals. On the one hand, it is a common
belief that numbers are abstractly represented (Dehaene, Dehaene-Lambertz, & Cohen,
1998) and hence other notations should yield effects similar to the current study. On the
other hand, it might be that the spatial representation is weaker in number words (at least
the automatic aspect of the spatial representation (Fias, 2001)), and the verbal code is
stronger due to the use of words, thus, giving rise to the effects that were observed here.
Future studies are needed to uncover the relationship between numbers and language in
general, and shared brain areas in particular.
Acknowledgements
This work was supported by the Kreitman Foundation (R.C.K.) and by the Israel Science Foundation founded by the Israel Academy of Sciences and Humanities (A.H.).
The authors are grateful to Ms. Desiree Meloul, Ms. Kathrin Cohen Kadosh for their
helpful suggestions, and the two anonymous reviewers for their very constructive comments.
References
Bachtold, D., Baumuller, M., & Brugger, P. (1998). Stimulus-response compatibility in representational space.
Neuropsychologia, 36, 731–735.
Berch, D. B., Foley, E. J., Hill, R. J., & Ryan, P. M. (1999). Extracting parity and magnitude from Arabic
numerals: developmental changes in number processing and mental representation. Journal of Experimental
Child Psychology, 74, 286–308.
380
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
Besner, D., & Coltheart, M. (1979). Ideographic and alphabetic processing in skilled reading of English.
Neuropsychologia, 17, 467–472.
Bloch-David, Y., Henik, A., & Rafal, R., submitted for publication. Intraparietal cortex and numerical distance.
Cohen Kadosh, R., & Henik, A. (2006a). Color congruity effect: where do colors and numbers interact in
synesthesia? Cortex, 42, 259–263.
Cohen Kadosh, R., & Henik, A. (2006b). When a line is a number: Color yields magnitude information in a digitcolor synesthete. Neuroscience, 137, 3–5.
Cohen Kadosh, R., & Henik, A. (2006c). A common representation for semantic and physical properties: a
cognitive-anatomical approach. Experimental Psychology, 53, 87–94.
Cohen Kadosh, R., Henik, A., & Rubinsten, O. (2005a). Are Arabic and number words processed in different
ways? Manuscript submitted for publication.
Cohen Kadosh, R., Henik, A., Rubinsten, O., Mohr, H., Dori, H., Van de Ven, V., et al. (2005b). Are
numbers special? The comparison systems of the human brain investigated by fMRI. Neuropsychologia,
43, 1238–1248.
Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed and stimulus-driven attention in the brain.
Nature Reviews Neuroscience, 3, 201–215.
Dehaene, S. (1996). The organization of brain activations in number comparison: event-related potentials and the
additive-factors method. Journal of Cognitive Neuroscience, 8, 47–68.
Dehaene, S., Bossini, S., & Giraux, P. (1993). The mental representation of parity and number magnitude.
Journal of Experimental Psychology: General, 122, 371–396.
Dehaene, S., Dehaene-Lambertz, G., & Cohen, L. (1998). Abstract representations of numbers in the animal and
human brain. Trends in Neurosciences, 21, 355–361.
Dehaene, S., Piazza, M., Pinel, P., & Cohen, L. (2003). Three parietal circuits for number processing. Cognitive
Neuropsychology, 20, 487–506.
Fias, W. (2001). Two routes for the processing of verbal numbers: evidence from the SNARC effect. Psychological
Research, 65, 242–249.
Fias, W., & Fischer, M. H. (2004). Spatial representation of numbers. In J. I. D. Campbell (Ed.). Handbook of
mathematical cognition (pp. 43–54). New York: Psychology Press.
Fias, W., Brysbaert, M., Geypens, F., & d’Ydewalle, G. (1996). The importance of magnitude information in
numerical processing: Evidence from the SNARC effect. Mathematical Cognition, 2, 95–110.
Fias, W., Lauwereyns, J., & Lammertyn, J. (2001). Irrelevant digits affect feature-based attention depending on
the overlap of neural circuits. Cognitive Brain Research, 12, 415–423.
Fias, W., Lammertyn, J., Reynvoet, B., Dupont, P., & Orban, G. A. (2003). Parietal representation of symbolic
and nonsymbolic magnitude. Journal of Cognitive Neuroscience, 15, 1–11.
Foltz, G. S., Poltrok, S. E., & Potts, G. R. (1984). Mental comparison of size and magnitude: Size congruity
effects. Journal of Experimental Psychology: Learning, Memory and Cognition, 10, 442–453.
Henik, A., & Tzelgov, J. (1982). Is three greater than five: the relation between physical and semantic size in
comparison tasks. Memory & Cognition, 10, 389–395.
Hubbard, E. M., Piazza, M., Pinel, P., & Dehaene, S. (2005). Interactions between number and space in parietal
cortex. Nature Reviews Neuroscience, 6, 435–448.
Kaufmann, L., Koppelstaetter, F., Delazer, M., Siedentopf, C., Rhomberg, P., Golaszewski, S., et al. (2005).
Neural correlates of distance and congruity effects in a numerical Stroop task: an event-related fMRI study.
Neuroimage, 25, 888–898.
Keppel, G. (1991). Design and analysis: a researcher’s handbook (3rd ed.). Upper Saddle River, NJ: Prentice
Hall.
Kinsbourne, M., & Hicks, R. E. (1978). Functional cerebral space: a model for overflow, transfer and interference
effects in human performance. A tutorial review. In J. Requin (Ed.). Attention and performance VII. Hillsdale,
New Jersey: Erlbaum.
Lammertyn, J., Fias, W., & Lauwereyns, J. (2002). Semantic influences on feature-based attention due to overlap
of neural circuits. Cortex, 38, 878–882.
Masson, M. E. J., & Loftus, G. R. (2003). Using confidence intervals for graphically based data interpretation.
Canadian Journal of Experimental Psychology, 57, 203–220.
Moyer, R. S., & Landauer, T. K. (1967). Time required for judgment of numerical inequality. Nature, 215,
1519–1520.
Naccache, L., & Dehaene, S. (2001). The priming method: imaging unconscious repetition priming reveals an
abstract representation of number in the parietal lobes. Cerebral Cortex, 11, 966–974.
R. Cohen Kadosh et al. / Acta Psychologica 124 (2007) 370–381
381
Pinel, P., Le Clec’h, G., van de Moortele, P.-F., Naccache, L., Le Bihan, D., & Dehaene, S. (1999). Even-related
fMRI analysis of the cerebral circuit for number comparison. NeuroReport, 10, 1473–1479.
Pinel, P., Dehaene, S., Rivie‘re, D., & Le Bihan, D. (2001). Modulation of parietal activation by semantic distance
in a number comparison task. NeuroImage, 14, 1013–1026.
Pinel, P., Piazza, M., Le Bihan, D., & Dehaene, S. (2004). Distributed and overlapping cerebral representations of
number, size, and luminance during comparative judgments. Neuron, 41, 983–993.
Posner, M. I., Sandson, J., Dhawan, M., & Shulman, G. L. (1990). Is word recognition automatic? A cognitiveanatomical approach. Journal of Cognitive Neuroscience, 1, 50–60.
Schwarz, W., & Ischebeck, A. (2000). Sequential effects in number comparison. Journal of Experimental
Psychology: Human Perception and Performance, 26, 1606–1621.
Schwarz, W., & Ischebeck, A. (2003). On the relative speed account of the number-size interference in
comparative judgment of numerals. Journal of Experimental Psychology: Human Perception and Performance,
29, 507–522.
Sternberg, S. (1969). The discovery of processing stages: Extension of Donders’ method. Acta Psychologica, 30,
276–315.
Tzelgov, J., Meyer, J., & Henik, A. (1992). Automatic and intentional processing of numerical information.
Journal of Experimental Psychology: Learning, Memory and Cognition, 18, 166–179.
Wiese, H. (2003). Iconic and non-iconic stages in number development: the role of language. Trends in Cognitive
Sciences, 7, 385–390.
Zorzi, M., Priftis, K., & Umilta, C. (2002). Neglect disrupts the mental number line. Nature, 417, 138–139.